J. Am. Chem. Soc. 1996, 118, 3069-3070
3069
Electrostatic Catalysis of the Claisen Rearrangement: Probing the Role of Glu78 in Bacillus subtilis Chorismate Mutase by Genetic Selection Peter Kast, Jeffrey D. Hartgerink, M. Asif-Ullah, and Donald Hilvert* The Scripps Research Institute Departments of Chemistry and Molecular Biology 10666 North Torrey Pines Road La Jolla, California 92037 ReceiVed NoVember 3, 1995
Figure 1. (A) Schematic view of the active site of wild-type BsCM2,3 with the presumed transition state for the chorismate-to-prephenate rearrangement (thin lines). (B) Electrostatic surface potential (negative, red; positive, blue) of the enzyme calculated with the program GRASP.21 A transition state analog is complexed at the active site, and the negatively charged Glu78 is visible as the red patch at the bottom of the predominantly positively charged binding pocket.
Conversion of (-)-chorismate to prephenate by chorismate mutase (CM) is a rare example of an enzyme-catalyzed sigmatropic rearrangement.1 Structural studies of CMs from Bacillus subtilis2,3 and Escherichia coli4 (BsCM and EcCM, respectively) indicate that the enzymes exert considerable conformational control over the flexible chorismate molecule.5 Both proteins also possess cationic and anionic residues oriented for potential electrostatic stabilization of a putative dipolar,6,7 chairlike8 transition state in which C-O bond cleavage substantially precedes C-C bond formation7,9 (Figure 1A). Theoretical studies have examined the importance of such electrostatic interactions,10 and mutagenesis experiments with BsCM have confirmed that a positive charge close to the ether oxygen of the substrate is essential for activity.11 We have now used targeted randomizing mutagenesis and selection to probe the role of Glu78. Our results support the notion3,12 that an electrostatic gradient in the active site is a major factor in CM catalysis. The chorismate-to-prephenate rearrangement is an essential step in the biosynthesis of phenylalanine and tyrosine in E. coli,1 making direct genetic selection for CM activity possible.11 To evaluate the role of key residues in the catalytic site of the aroHencoded BsCM,13 the corresponding codons can be randomized and functional variants of the protein selected for their ability to allow a host strain lacking endogenous CM genes to grow on minimal medium.11 As shown in Figure 1, the position of Glu78 at the bottom of the binding pocket suggests3,12 that its carboxylate could stabilize the partial positive charge on the cyclohexadienyl ring of the presumed dipolar transition state. To assess the importance of a negative charge at position 78,
aroH libraries I (randomized codon 78) and II (simultaneously randomized positions 75 and 78) were constructed.14 The libraries were introduced into the host strain KA12/ pKIMP-UAUC,11 and clones able to produce their own phenylalanine and/or tyrosine were analyzed by sequencing.11 Additionally, randomly chosen control clones grown under nonselective conditions were included.15 Figure 2A,B compiles all unambiguously identified CM variants (without secondary mutations or wild-type codons) from libraries I and II, respectively. Table 1 summarizes additional in ViVo and in Vitro data obtained with a subset of clones.16 Intracellular concentrations of CM were shown by SDS-polyacrylamide gel electrophoresis to vary by less than a factor of 2 for all mutants except for a clone with an amber stop codon at position 78, where no CM band was visible (data not shown). Strain KA12 (supE44) partially suppresses this stop codon with Gln.17 When only codon 78 was randomized, all 20 fast-growing clones encoded Glu78 (Figure 2A). The fact that the isosteric but uncharged variant Gln78 is essentially inactive (Table 1) underscores the importance of the negative charge of Glu78. That the carboxylate provides more than a hydrogen bond to chorismate’s hydroxyl group is consistent with substrate modification experiments showing that the substrate alcohol is not essential for catalysis by an EcCM.18 Nevertheless, a low level of activity is obtained when Glu78 is replaced with some residues capable of hydrogen bonding, such as His, Cys, Ser,
(1) Haslam, E. Shikimic Acid: Metabolism and Metabolites; John Wiley & Sons: New York, 1993. (2) Chook, Y. M.; Ke, H.; Lipscomb, W. N. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 8600-8603. (3) Chook, Y. M.; Gray, J. V.; Ke, H.; Lipscomb, W. N. J. Mol. Biol. 1994, 240, 476-500. (4) Lee, A. Y.; Karplus, P. A.; Ganem, B.; Clardy, J. J. Am. Chem. Soc. 1995, 117, 3627-3628. (5) Lee, A. Y.; Stewart, J. D.; Clardy, J.; Ganem, B. Chem. Biol. 1995, 2, 195-203. (6) Copley, S. D.; Knowles, J. R. J. Am. Chem. Soc. 1987, 109, 50085013. (7) Gajewski, J. J.; Jurayj, J.; Kimbrough, D. R.; Gande, M. E.; Ganem, B.; Carpenter, B. K. J. Am. Chem. Soc. 1987, 109, 1170-1186. (8) Sogo, S. G.; Widlanski, T. S.; Hoare, J. H.; Grimshaw, C. E.; Berchtold, G. A.; Knowles, J. R. J. Am. Chem. Soc. 1984, 106, 27012703. (9) Addadi, L.; Jaffe, E. K.; Knowles, J. R. Biochemistry 1983, 22, 44944501. (10) (a) Lyne, P. D.; Mulholland, A. J.; Richards, W. G. J. Am. Chem. Soc. 1995, 117, 11345-11350. (b) Wiest, O.; Houk, K. N. J. Am. Chem. Soc. 1995, 117, 11628-11639. (11) Kast, P.; Asif-Ullah, M.; Jiang, N.; Hilvert, D. Proc. Natl. Acad. Sci. U.S.A., in press. (12) Haynes, M. R.; Stura, E. A.; Hilvert, D.; Wilson, I. A. Science 1994, 263, 646-652. (13) Gray, J. V.; Golinelli-Pimpaneau, B.; Knowles, J. R. Biochemistry 1990, 29, 376-383.
(14) The partially randomized aroH libraries I and II were constructed using the polymerase chain reaction (PCR) with oligonucleotides 5′ATCAAGCGGCCGCACTAGT and 5'-GTATGTACCGGTAACATGTATGCAGNN(G/C)ATGGACGTCACAGGCGGTCTTA (library I) or 5′-GTATGTACCGGTTACTNN(G/C)ATGCAGNN(G/C)ATGGATGTCACAGGCGGTCTTA (library II). Randomized codons are italicized. The gel-purified 197 base pair (bp) PCR products were digested with PinAI (site underlined) and HindIII. The 144 bp library fragments were ligated into a 3017 bp PinAI-HindIII fragment of plasmid pKCMT-W previously incubated with BsrGI and gel purified. Detailed procedures and subsequent manipulations are described in ref 11; wild-type aroH contamination was eliminated by digestion of the library II plasmid pool with AatII prior to transformation of the selection strain KA12/pKIMP-UAUC. (15) Sequencing of control clones revealed that codon 78 in library I exhibited a biased (4-fold lower) T-content, probably due to unequal coupling efficiencies during oligonucleotide synthesis. Nevertheless, we believe that the large number of 51 selected active clones surveyed gives a representative picture of permissible alternatives at position 78. The composition of library II was random. (16) In general, there was good correlation between assays in ViVo (growth on plates and doubling time determinations) and in Vitro (enzyme activity). Exceptions can probably be explained by instability of individual variants in Vitro as a consequence of the location of the alterations at the subunit interface. Comparisons between the assays used are discussed in ref 11. (17) Eggertsson, G.; So¨ll, D. Microbiol. ReV. 1988, 52, 354-374. (18) Pawlak, J. L.; Padykula, R. E.; Kronis, J. D.; Aleksejczyk, R. A.; Berchtold, G. A. J. Am. Chem. Soc. 1989, 111, 3374-3381.
0002-7863/96/1518-3069$12.00/0
© 1996 American Chemical Society
3070 J. Am. Chem. Soc., Vol. 118, No. 12, 1996
Communications to the Editor Table 1. Characterization of a Subset of Selected Mutants clone wild typec vectorc His78 Cys78 Ser78 Asn78 Gly78 Pro78 Gln78 amber78 Ala78 Met78 Asp78Gly81 Ser75Glu78 Asp75Glu78 Gly75Glu78 Met75Glu78 Val75Glu78 Asp75Ser78 Asp75Ala78 Ser75Asp78 Asp75Met78 Val75Ser78 Asp75Val78 Cys75Ser78 Ile75Asn78
Figure 2. Summary of results from analysis of libraries randomized at position 78 (A) and positions 75 and 78 (B) of BsCM. The vertical axis displays in ViVo activity (colony size on selective media) of individual clones ranging from 0 (no growth) to 5 (fast, wild-type-like growth).11 Numbers on the tops of columns indicate the occurrence of that variant. Clones encoding Glu78 are identified by red columns. Amino acids are ordered according to increasing side-chain volume.22
or Asn (Table 1), perhaps because of compensatory structural changes at the active site. The importance of the negative charge is further emphasized by the pattern of residues selected from the combinatorial library II. As visualized in Figure 2B, the only variants with wildtype-like activity in ViVo contain either Glu at position 78 or Asp at position 75. The relatively high activities in ViVo and in Vitro of variants Ser75Glu78 and Gly75Glu78 (Table 1) show that Cys at position 75 is unessential, contrary to recent speculations.5 A carboxylate group alone is not sufficient for complementation, however; its proper placement appears equally important,19 given that Asp78 only appears in selected library I clones as the fortuitous double-mutant Asp78Gly81.16 The Val81-to-Gly81 exchange may cause a surface loop to pack (19) For analogous findings in other enzymes, see: (a) Hibler, D. W.; Stolowich, N. J.; Reynolds, M. A.; Gerlt, J. A.; Wilde, J. A.; Bolton, P. H. Biochemistry 1987, 26, 6278-6286. (b) Hermes, J. D.; Blacklow, S. C.; Knowles, J. R. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 696-700. (20) Hilvert, D.; Carpenter, S. H.; Nared, K. D.; Auditor, M.-T. M. Proc. Natl. Acad. Sci. U.S.A. 1988, 85, 4953-4955. (21) Nicholls, A.; Sharp, K. A.; Honig, B. Proteins 1991, 11, 281-296. (22) Chothia, C. Annu. ReV. Biochem. 1984, 53, 537-572.
phenotype in ViVoa on plates td (h) 5 0 3 2 2 2 0.5 0.5 0 0 0 0 5 5 5 5 5 5 5 5 4.5 4 3 3 2 1.5
2.8 no growth 4.5 nt 7.7 7.9 9.0 15 9.9 nt 17 nt 3.1 2.5 nt nt nt nt 3.0 nt 3.4 4.7 7.4 nt nt nt
specific activity in Vitrob (% of wild type) 100
